The present invention relates generally to compounds, termed ligands, which bind to metals in a very stable manner. Specifically, it relates to ligands which bind to metals with sufficient stability that the resulting ligand-metal complexes can remain intact in vivo. Such complexes find a variety of uses in medicine, of particular interest to this invention being their use in radiopharmaceutical applications when the metal atom is a radiation emitting or absorbing isotope. According to certain embodiments, the invention relates to bifunctional ligand molecules capable of binding metals to substrates such as antibodies and other proteins by means of a specific substrate reactive group having a phenyl ring which is meta or para substituted with a substrate reactive moiety. Antibody-metal ion conjugates may be produced for use in vivo diagnostic imaging methods as well as in therapeutic methods where the metal ions emit cytotoxic radiation.
Many factors influence the stability of ligand-metal complexes, including both characteristics inherent in the metal itself and characteristics related to the molecular structure of the ligand molecule. The properties of metal atoms, while they can vary widely from metal to metal, are generally not amenable to premeditated manipulations intended to enhance complex stability, with the result that ligand structure is the focus of efforts in the art to maximize the stability of such complexes. Ligand molecules known to the art and relevant to this invention are typically organic molecules of relatively low molecular weight (100-2,000 daltons) which may be either naturally occurring compounds or synthetic materials subject to premeditated design.
Certain design criteria for producing useful synthetic ligands are well known in the art. The nature of the chemical bond formed between a ligand and a metal can be thought of as involving a donation of a pair of electrons present on the ligand molecule or, in molecular orbital terms, as a molecular orbital formed by combining a filled orbital on the ligand with a vacant orbital on the metal atom. Those atoms in the ligand molecule which are directly involved in forming a chemical bond to the metal atom are thus termed the donor atoms and these generally comprise elements of Groups V and VI of the periodic table with nitrogen, oxygen, sulfur, phosphorus and arsenic being those most commonly employed.
An important criterion in designing synthetic ligand molecules is based on the finding that the stability of a metal complex formed with a ligand containing two donor atoms within the same molecule is generally greater than that of the corresponding complex formed with two separate ligand molecules, each containing only a single donor atom, even though the metal and the donor atoms are the same in each case. A second important design criterion, is based on the limitation that the stereochemistry of the ligand molecule must be such as to permit both donor atoms to bind to the metal atom without producing severe steric strain in the framework of the ligand molecule.
Ligand molecules that contain two or more donor atoms capable of binding to a single metal atom are termed chelating agents or chelators and the corresponding metal complexes are called chelates. The underlying thermodynamic phenomenon responsible for the enhanced stability obtained when two or more donor atoms are present in the same ligand molecule is called the chelate effect, as described in detail in F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, 4th edition, pp. 71-73, published in 1980. The number of donor atoms present in a given chelator is termed the denticity of that chelator, ligands possessing two donor sites being called bidentate, those with three donor sites, tridentate, and so forth. Although the foregoing example used the case of a bidentate ligand to illustrate the chelate effect, the effect also applies to ligands of higher denticity, up to the point at which the denticity of the chelator matches the maximum coordination number attainable by the particular metal atom.
The maximum coordination number of a metal atom is an intrinsic property of that atom, relating to the number of vacant orbitals that the atom possesses and hence the number of chemical bonds it is able to form. For a given metal atom in a given oxidation state, the maximum coordination number is largely invariant and, as a rule, cannot be altered by changing the nature of the ligands. When all of the available vacant orbitals have been used to form bonds to donor atoms in the ligand or ligands, the metal atom is said to be coordinatively saturated. In general, the maximum stability of a metal complex is thus achieved when the complex is formed using a multidentate chelator, the denticity of which is sufficient to coordinatively saturate the metal atom. It is therefore an object of this invention to provide multidentate chelators capable of producing coordinatively saturated complexes of radiopharmaceutically useful metals.
Coordinative saturation for most metals of the first and second transition series and for the group IIIb metals of particular interest to the present invention is generally achieved at a coordination number of six. Attempts in the art to design chelators which would form maximally stable complexes with these metals have thus centered on the development of ligands containing six donor atoms. Of particular interest to the art has been the development of hexadentate chelators forming highly stable complexes with the biologically important metal iron(III). These efforts have often involved a biomimetic approach, since the most stable iron(III) complexes presently known are those formed by naturally occurring chelators produced by a variety of microorganisms as part of their mechanism for obtaining iron from the environment. Such chelators are termed siderophores, the most stable of all metal complexes known to the art being the iron(III) complex of a siderophore called enterobactin, which has the structure: ##STR1##
The formation constant of the ironenterobactin complex is 10.sup.52, as reported by Harris, et al., J. Amer. Chem. Soc., 101, 6097 (1979). Enterobactin binds to iron through the six phenolic oxygen atoms present on the three catechol moieties. Thus, the overall chelate structure is composed of three identical bidentate binding units, each unit comprising the structure ##STR2## The remainder of the molecule functions as a framework which positions the three bidentate catechol moieties such that all three are stereochemically capable of binding to a single iron center.
A second powerful siderophore, which is used clinically to induce iron excretion, is called desferrioxamine B and has the structure ##STR3## This chelator binds to iron through an array of six oxygen atoms present on three hydroxamate bidentate binding units, each unit having the structure: ##STR4## As is the case with enterobactin, the remainder of the molecule serves as a framework possessing sufficient steric flexibility to permit all three bidentate hydroxamate moieties to bind to the same iron atom.
Attempts in the art to mimic the foregoing siderophore structures have focused on the production of compounds employing synthetic frameworks to position the natural catecholate and hydroxamate moieties appropriately for formation of hexadentate chelates.
Harris, et al., disclose a synthetic hexadentate chelating compound 1,3,5-N,N',N"-tris(2,3-dihydroxybenzoyl)triamomethyl benzene (MECAM) wherein three catechol moieties are linked to a benzene backbone moiety.
Jain, et al., J. Amer. Chem. Soc., 89, 724 (1967) and J. Amer. Chem. Soc., 90, 519 (1968) disclose the use of the trifurcate tetraamine tris-(2-aminoethyl)amine (TREN) as a tetradentate chelator for metals such as copper and zinc.
Rodgers, et al., Inorg. Chem., 26, 1622 (1987) and references therein disclose the synthesis of a number of hexadentate tricatechol analogs of enterobactin including one utilizing TREN as a backbone. The compounds form very stable complexes with iron.
Weitl et al., U.S. Pat. Nos. 4,181,654, 4,309,305 and 4,543,213 disclose compounds comprising four 2,3-dihydroxybenzoic acid amide (catechol) moieties arranged about a cyclic or linear azaalkane framework. The dihydroxybenzoyl groups may optionally be substituted with a nitro group, a sulfonate group or a pharmaceutically acceptable sulfonate salt. The compounds are disclosed to be particularly useful for sequestering actinide (IV) ions by formation of octadentate complexes. Weitl, et al., U.S. Pat. No. 4,442,305 relates to polybenzamide compounds of lower denticity comprising two 2,3-dihydroxy benzoyl groups linked by an azaalkane framework. Such quatradentate compounds are also said to be useful for sequestering actinide (IV) ions according to both in vivo and in vitro procedures.
Of interest to the present invention is art relating to chelating agents utilizing atoms other than oxygen as electron donors. The polyaminopolycarboxylate chelators such as ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA), which are well known in the art and form stable complexes with a wide range of metals, derive much of their utility from the fact that they present a combination of both oxygen and nitrogen donors. The synthesis of bifunctional DTPA and EDTA analogues is well known. Sundberg, et al., J. Med. Chem., 17, 1304 (1974) discloses the synthesis of bifunctional EDTA derivatives characterized by the attachment of unique protein substrate reactive functions such as paraaminophenyl protein reactive substituents at a methylene carbon of the polyamine backbone.
Of interest to the present invention are disclosures relating to 8-hydroxyquinoline, also known as oxine, a bidentate binding unit containing both an oxygen donor atom and a nitrogen donor atom which has the structure shown below when bound to a metal atom, M ##STR5## As is well known in the art, 8-hydroxyquinoline forms complexes with a wide range of metals and is known for use in extraction of various metal ions from solutions. See, Plueddemann, U.S. Pat. No. 4,421,654 and Scher, U.S. Pat. No. 4,500,494, which discloses a variety of substituted oxines. 8-hydroxyquinoline is known to be particularly useful for complexing group IIIb metals. The latter are of particular interest to the present invention since the heavier group IIIb elements find wide use in nuclear medicine, especially the gamma emitting isotopes gallium-67, indium-111 and thallium-201 and the positron emitting metal gallium-68. McAfee, et al., J. Nucl. Med., 17, 480 (1976) discloses the labelling of blood cells with 8-hydroxyquinoline complexes of indium-111 and technetium-99m. Moerlein, et al., Int. J. Nucl. Med. Biol., 8, 277 (1981), discloses various linear and branched 2,3-dihydroxybenzoylamide analogs of enterobactin and their use with both gallium and indium as radiopharmaceuticals. The reference also discloses the use of indium-111 labelled 8-hydroxyquinoline. Loberg, et al., U.S. Pat. No. 4,017,596 disclosed the use of chelates of cobalt-57, gallium-67, gallium-68, technetium-99m, indium-111 and indium-113m with 8-hydroxyquinoline as radiopharmaceutical external imaging agents. Goedemans, et al., European Patent Application No. 83,129 discloses antibodies and antibody fragments labelled with radionuclides through bifunctional chelating agents including 8-hydroxyquinoline. The chelating agents of the references form complexes in which three 8-hydroxyquinoline ligands are bound to each metal atom.
While the tris(8-hydroxyquinoline) complexes are satisfactory for the foregoing applications, they are of limited utility as radiopharmaceuticals for in vivo use as the bidentate 8-hydroxyquinoline ligand has a formation constant for, for example, gallium of only 10.sup.14.5 whereas that of the serum protein transferrin is 10.sup.23.7. Consequently, once in the bloodstream, such complexes would tend to break down and release the radioactive metal to transferrin, resulting in undesirably high and persistent blood background levels of radioactivity.
Hata, et al., Bull. Chem. Soc. Japan, 45, 477 (1972) discloses azomethyl and azomethine derivatives of 8-hydroxyquinoline-2-carbaldehyde. One derivative is N,N'-Bis(8-hydroxy-2 quinolylmethyl)ethylene diamine comprising two 8-hydroxyquinoline chelating moieties attached to an ethylene diamine framework. The reference suggests that the tetradentate chelator may form more stable metal complexes than oxine as a consequence of its higher basicity.
Of interest to the present invention are disclosures showing the use of protein/metal ion conjugates for diagnostic and therapeutic purposes. Gansow, et al., U.S. Pat. No. 4,454,106 discloses the use of monoclonal antibody/metal ion conjugates for in vivo and in vitro radioimaging diagnostic methods. Goldenberg, et al., N. Eng. J. Med., 298, 1384-88 (1978) discloses diagnostic imaging experiments wherein antibodies to the known tumor associated antigen carcinoembryonic antigen (CEA) are labelled with 131iodine and injected into patients with cancer. After 48 hours, the patients are scanned with a gamma scintillation camera and tumors are localized by the gamma emission pattern.
Other workers disclose the therapeutic use of antibody/metal ion conjugates for delivery of cytotoxic radioisotopes to tumor deposits in vivo. Order, et al., Int. J. Radiation Oncology Biol. Phys., 12, 277-81 (1986) describes treatment of hepatocellular cancer with antiferritin polyclonal antibodies to which 90yttrium has been chelated. Buchsbaum, et al., Int. J. Nucl. Med. Biol., Vol. 12, No. 2, pp. 79-82, (1985) discloses radiolabelling of monoclonal antibodies to CEA with .sup.88 yttrium and suggests the possibility of localization and treatment of colorectal cancers therewith. Nicolotti, EPO Application No. 174,853 published March 19, 1986, discloses conjugates comprising metal ions and antibody fragments. According to that disclosure, monoclonal antibodies of subclass IgG are enzymatically treated to remove the Fc fragment and reductively cleave the disulfide bond linking the antibody heavy chains. The Fab' fragment is then linked to a chelating agent bound to a radionuclide metal ion for in vivo diagnostic or therapeutic use. Also of interest to the present application is the disclosure of Wang, et al., J. Nucl. Med., 28, 723 (1987), which relates to a bifunctional bidentate chelator derived from the naturally occurring 8-hydroxyquinoline derivative xanthurenic acid. This chelator has the structure ##STR6## The carboxylic acid is converted to an N-hydroxysuccinimidyl ester, which reacts with the side chain amino groups of proteins to link the bidentate chelator to the substrate. The stability of a single bidentate unit tends to be insufficient for in vivo applications and the use of a complex containing three xanthurenic acid active ester ligands would generally be expected to lead to deleterious cross linking between proteins and concomitant denaturation of the protein substrate.